bartwr

10 ms·

91.
by bartwr 6y ago
So I don't know how HN really works, but FWIW I posted it here 10mins after the post went live. Then it just went down into obscurity (almost always my submissions to HN do), so I just ignored it, later saw a repost and ignored it, &qu
92.
by bartwr 6y ago
I mention the Dreams in the passing - as it is such a special gem and special case that it would require much more care and analysis, probably material for multiple posts. What makes Dreams special is that it is a platform for creativity an
93.
by bartwr 6y ago
In the article, I mention a few ways of dealing with it - from "old school" impostors/billboards (think of precomputed sprites), simplified representation, to finally temporal anti-aliasing (reusing pixels in screen space). A
94.
Why are video games graphics challenging? Productionizing rendering algorithms(bartwronski.com)
309 points by bartwr 6y ago|89 comments
95.
Compressing PBR material texture sets with sparsity and kSVD dictionary learning(bartwronski.com)
2 points by bartwr 6y ago|0 comments
96.
Dimensionality reduction for image and texture set compression(bartwronski.com)
3 points by bartwr 6y ago|0 comments
97.
“Optimizing” blue noise dithering – backpropagation through Fourier and sorting(bartwronski.com)
3 points by bartwr 6y ago|0 comments
98.
Bilinear texture filtering – artifacts, alternatives, frequency domain analysis(bartwronski.com)
1 points by bartwr 6y ago|0 comments
99.
Using JAX, numpy, and optimization techniques to improve separable image filters(bartwronski.com)
101 points by bartwr 7y ago|9 comments
100.
Separability, SVD, and low-rank approximation of 2D image processing filters(bartwronski.com)
1 points by bartwr 7y ago|0 comments
101.
Analyze your own activity data using Google Takeout – music stats example(bartwronski.com)
1 points by bartwr 7y ago|0 comments
102.
by bartwr 7y ago
Hey, I'm Bart, author of the blog post that got linked in this HN. :) Your points re error diffusion being way better are true; the difference is night-and-day, especially on some thin features / lines. At the same time, error di